Phylogeny of the Aplousobranchia (Tunicata: Ascidiacea) 1 Tatiane R. Moreno 2 & Rosana M. Rocha 2 1 Contribuition number 1754 of the Departamento de Zoologia, Universidade Federal do Paraná. 2 Laboratório de Biologia e Ecologia de Ascidiacea, Departamento de Zoologia, Universidade Federal do Paraná. Caixa Postal 19020, 81531-980 Curitiba, Paraná, Brasil. E-mail: [email protected]; [email protected] ABSTRACT. The phylogenetic relationships of genera and families of Aplousobranchia Lahille (Tunicata, Ascidiacea) is reconstructed based on morphological characters – the first comprehensive morphology-based phylogenetic analysis for the Aplousobranchia. Monophyly of Aplousobranchia and its families were tested with samples of 14 families. The final character matrix comprised 47 characters and 41 genera as terminal taxa. Nine equally most parsimonious trees (length 161, CI = 0.5031, RI = 0.7922) were found. Characters describing replication, colony system formation, and branchial walls were the more important in phylogenetic reconstruction. These characters were more useful than others more traditionally used in ascidian taxonomy, such as: body division, position of the heart, gonads and epicardium. Characters not frequently used in phylogenetic analysis, such as body wall muscles, muscles associated with transversal blood vessels and arrangement of the larval papillae, also have phylogenetic information. Results supported monophyly of the Aplousobranchia sensu Lahille, 1887 including only Polycitoridae, Polyclinidae, and Didemnidae. On the other hand, Aplousobranchia including also Cionidae and Diazonidae is not monophyletic since Perophora and Ecteinascidia were included as ingroups in the cladogram, Ciona (now closer to Ascidia) was no longer included in Aplousobranchia and the position of Rhopalaea and Diazona is not resolved. We propose a revised classification based on this phylogenetic analysis, in which Aplousobranchia, with three new families and an indeterminate taxon, now has 15 families. KEY WORDS. Chordata; classification; morphological characters; systematic; Urochordata. RESUMO. Filogenia de Aplousobranchia (Tunicata: Ascidiacea). O relacionamento filogenéticos de gêneros e famílias de de Aplousobranchia Lahille (Tunicata, Ascidiacea) foi reconstruída com base em caracteres morfológicos – esta constitui a primeira análise filogenética morfológica abrangente para Aplousobranchia. A monofilia de Aplousobranchia e suas famílias foi testada com espécies de 14 famílias. A matriz final compreendeu 47 caracteres e 41 gêneros como taxons terminais. Nove árvores igualmente parcimoniosas (comprimento 161 passos, CI = 0.5031, RI = 0.7922) foram encontradas. Caracteres descrevendo reprodução, formação de sistemas de zoóides nas colônias e relacionados com a parede da faringe foram os mais importantes na reconstrução filogenética. Estes caracteres apresentaram maior sinal filogenético do que outros mais tradicionalmente utilizados na taxonomia de Ascidiacea como divisão do corpo, posição do coração, gônadas e epicárdio. Caracteres pouco utilizados como musculatura da parede do corpo, musculatura associada aos vasos sanguíneos transversais da faringe e arranjo das papilas adesivas da larva também apresentaram informação filogenética. Os resultados suportam a monofilia de Aplousobranchia sensu Lahille, 1887 incluindo apenas Polycitoridae, Polyclinidae e Didemnidae. Por outro lado, Aplousobranchia incluindo também Cionidae e Diazonidae não é taxon monofilético já que Perophora e Ecteinascidia foram incluídos como grupos internos no cladograma; Ciona (próximo a Ascidia) não pôde ser incluído em Aplousobranchia e não foi possível definir a posição de Rhopalaea e Diazona. Uma classificação revisada baseada nesta análise filogenética é proposta, na qual Aplousobranchia possui 15 famílias, sendo três novas famílias e um táxon indeterminado. PALAVRAS-CHAVE. Caracteres morfológicos; Chordata; classificação; sistemática; Urochordata. Traditionally, pharynx structure provided the basis for as- pharynx, with numerous internal longitudinal vessels); and cidian classification (LAHILLE 1887, 1890 in KOTT 1969) while later Stolidobranchia (folded pharynx with internal longitudinal ves- studies included the position of the gonads (PERRIER 1898 in KOTT sels). Using gonad position, the class is divided into two Orders: 1969). Based on the structure and development of the pharynx, Enterogona, with the gonads within or below the intestinal loop the class is divided into three Orders: Aplousobranchia (flat phar- (includes Aplousobranchia and Phlebobranchia); and Pleurogona, ynx without internal longitudinal vessels); Phlebobranchia (flat with gonads attached to the body wall (includes Stolidobranchia). Revista Brasileira de Zoologia 25 (2): 269–298, June, 2008 270 T. R. Moreno & R. M. Rocha BERRILL (1936) traced a probable ascidian evolutionary path- Very few attempts to reconstruct ascidian phylogeny ex- way based on the development of the heart, pericardium and ist, and the first ones were intuitive and not cladistically based. epicardium. He suggested that classification based on the phar- One of the first comprehensive attempts proposed the origin ynx was weak since pharyngeal structure is probably largely due of a group including Diazona, Rhopalaea, Tylobranchion and the to the size of the organism, and consequently divided the evolution of Aplousobranchia from a Ciona-like ancestor (BERRILL Ascidiacea into Pleurogona and Enterogona. More recently, HIROSE 1936). Following this hypothesis, Didemnidae would have a (2001), in a study of tunic cell types in 65 ascidians species (in- common origin with Distaplia while Perophoridae is an iso- cluding 11 of the 15 traditional ascidian families), showed that lated branch originating directly from Ciona. In the tree pro- two tunic cell types (“tunic bladder cells” and “tunic net cells”) duced by this hypothesis, Phlebobranchia was clearly polyphyl- supported the Pleurogona and Enterogona division, because they etic. KOTT (1962) proposed an alternative that suggested inde- were only found in the Enterogona. Additional analysis, based pendent origins for Polycitoridae, Polyclinidae and Didemnidae, on 18S rDNA sequences (WADA et al. 1992, WADA 1998), also sup- each with a Diazonidae-like ancestor, and which would make ported this view. In summary, two systems of classification were Aplousobranchia polyphyletic. MILLAR (1966) briefly discussed prominent in the twentieth century: 1) the orders Pleurogona the main lines of ascidian evolution and proposed a phyloge- and Enterogona and sub-orders Aplousobranchia, Phlebobran- netic tree based exclusively on morphological characters. He chia and Stolidobranchia (BERRILL 1936, 1950, KOTT 1985); 2) the suggested that Clavelinidae might have arisen close to the di- Aplousobranchia, Phlebobranchia and Stolidobranchia as orders vision between Aplousobranchia and Phlebobranchia. (VAN NAME 1945, MONNIOT & MONNIOT 1972, MONNIOT et al. 1991). Apart from these few morphological studies, some phylo- Monophyly of these groups must be determined to better choose genetic hypotheses based on specific characters were proposed between the two classifications. Here, we test the monophyly of in various studies. For example, vanadium as a character was the Aplousobranchia. proposed by HAWKINS et al. (1983). They noted that all the spe- The Order Aplousobranchia comprises colonial ascidians, cies of Aplousobranchia sensu Lahille and of the families Cionidae with the body divided into two or three parts (thorax, abdo- and Diazonidae contained significant concentrations of vana- men and, in some cases, posterior abdomen); with simple oral dium in its oxidized form (oxidation state IV). This character tentacles and the pharynx with neither folds nor internal lon- confirms the proximity of those groups and suggests that they gitudinal vessels (VAN NAME 1945). In a comprehensive review share a common ancestor (KOTT 1990), but not necessarily that of the ascidian genera of the world, MONNIOT & MONNIOT (1972) they form a monophyletic group. In all the Phlebobranchia, recognized 46 genera of Aplousobranchia in only three fami- vanadium appears in oxidation state III while, in Stolidobranchia, lies: Polyclinidae, Polycitoridae and Didemnidae. Two addi- significant levels of vanadium were rarely encountered. Trends tional families, Clavelinidae and Holozoidae, were later accepted in the evolution within each of those taxa seem to be associated (MONNIOT & MONNIOT 2001). with the loss of the element or changes of the oxidation state of KOTT (1969, 1985, 1990, 1992, 2001), based on long-term vanadium (IV for III). surveys of ascidians from Australia and Antarctica, proposed the Recently, accumulating molecular data have shed new light most important modifications in the traditional classification of on the phylogenetic relationships within Deuterostomia the Aplousobranchia: inclusion of Diazonidae, with the genera (CAMERON et al. 2000, TURBEVILLE et al. 1994, SWALLA et al. 2000, Rhopalaea Philippi, 1843, Diazona Savigny, 1816, Tylobranchion SWALLA 2001) and within Ascidiacea (WADA et al. 1992, WADA 1998, Herdman, 1886, Syndiazona Oka, 1926 and Pseudodiazona Millar, STACH & TURBEVILLE 2002, TURON & LÓPEZ-LEGENTIL 2004). Analysis 1963a; and inclusion of Cionidae, with Ciona Fleming, 1822. Those based on 18S rDNA sequences (WADA et al. 1992, WADA 1998) families belonged previously within the Phlebobranchia Lahille, inferred that Enterogona (including only Phlebobranchia) is 1887. She also proposed